[R] Roxygen vs Sweave for S4 documentation

[Duncan Murdoch]

On 6/22/2009 9:23 AM, Tobias Verbeke wrote:
> On Mon, Jun 22, 2009 at 2:18 PM, Douglas Bates<bates at stat.wisc.edu> wrote:
>> On Mon, Jun 22, 2009 at 7:12 AM, Martin
>> Maechler<maechler at stat.math.ethz.ch> wrote:
>>>>>>>> "TobiasV" == Tobias Verbeke <tobias.verbeke at openanalytics.be>
>>>>>>>>     on Sun, 21 Jun 2009 08:25:07 +0200 writes:
>>>
>>>    TobiasV> Hi Ken,
>>>    >> I have been using R for a while.  Recently, I have begun converting my
>>>    >> package into S4 classes.  I was previously using Rdoc for documentation.
>>>    >> Now, I am looking to use the best tool for S4 documentation.  It seems that
>>>    >> the best choices for me are Roxygen and Sweave (I am fine with tex).
>>>    >>
>>>    >> Are there any users of Roxygen or Sweave who can comment on the strengths or
>>>    >> weaknesses of one or othe other?  Thanks in advance.
>>>
>>>    TobiasV> For the moment proper documentation of S4 classes (with a @slot tag
>>>    TobiasV> e.g.) is not implemented yet,
>>>
>>> how did you define "proper" here?
>>>
>>> I know that the result of  promptClass()  may not always be
>>> perfect. As most things are not perfect,
>>> I would not quickly call this improper ...
>>>
>>> Or is using *.Rd  not proper for you,
>>> as it does not have all of code + docs in one file?
>>
>> I think Tobias was referring to Roxygen support for S4 classes, not
>> the existence of .Rd tags.
> 
> Yes, I was just referring to the Roxygen support and as a matter of
> fact I recommended Roxygen to mimick the nice job done by promptClass
> a while ago
> 
> https://lists.r-forge.r-project.org/pipermail/roxygen-devel/2009-February/000023.html
> 
> No intent to criticize nor current Rd tags nor Rd as a documentation system,
> although having code + docs in one file (as in Roxygen) seems to be convenient
> to some.

Hopefully the Rd changes coming in R 2.10.0 will enhance either way of 
generating an Rd file.  There are some things (e.g. all the methods for 
a generic, or all the generics with methods for a class, all the classes 
extending a given one) that aren't known at the time you run promptClass 
or Roxygen, but which are known at display time; the new system should 
allow those to be computed just before displaying the page.

Duncan Murdoch