[Roxygen-devel] roxygen3

[Andrew Redd]

I would be happy to contribute an extractor function to get the relevant 
comments.  We would have a dependency on parser or R>2.16 if we use 
Duncan's version that made it into the core-R. But we have all the 
machinery to extract the information.  Perhaps if we could form a 
standard for organizing the information so that the roccers could handle it.

-Andrew



On 08/28/2012 09:17 AM, Hadley Wickham wrote:
> Hi Andrew,
>
> The comments are still attached only to top-level expressions, so it
> still won't work.  I'm thinking about how to change that to deal with
> (e.g.) reference classes, so it's possible I'll figure it out, but I
> wouldn't hold your breath.
>
> Hadley
>
> On Tue, Aug 28, 2012 at 10:13 AM, Andrew Redd <amredd at gmail.com> wrote:
>> Hadley,
>> Great news.  Will we be able to do the Doxygen style commenting of function
>> parameters that I have discussed with you earlier?  Will that be possible or
>> should I still pursue doing that within lint?
>>
>> Thanks,
>> Andrew Redd
>>
>>
>> On 08/27/2012 05:53 PM, Hadley Wickham wrote:
>>> Hi all,
>>>
>>> I thought I should mention I've started working on roxygen3 at
>>> https://github.com/hadley/roxygen3.  It's a ground up rewrite of
>>> roxygen2, aiming to produce the same results as roxygen3, but without
>>> a completely different backend.  Currently the roxygen2 code is hard
>>> to extend - the idea of roclets was good, but I think they were too
>>> big - you want to be able to work at the tag level so it's easier to
>>> add new features.
>>>
>>> The new roccers in roxygen3 look something like this:
>>>
>>> parse_dev <- function(roc, ...) {
>>>     if (is.null(roc$dev)) return()
>>>     list(
>>>       title = str_c("[DEV] ", roc$title),
>>>       description = c("This function is useful only for developers",
>>>         roc$description),
>>>       dev = NULL)
>>> }
>>>
>>> add_roccer("dev", roc_parser(one = parse_dev))
>>> base_prereqs[["dev"]] <- c("_intro", "title", "details")
>>>
>>> That defines @dev which adds a note to the title and the description
>>> that the function is more suitable for developers than end-users.
>>>
>>> The interface is likely to change a lot, and I doubt the package
>>> currently installs (although it does work with devtools::load_all),
>>> but if you're interested please take a look. I'm hoping a stronger
>>> foundation will make it easier to keep on top of the bugs and to
>>> flexibly implement new features.
>>>
>>> Hadley
>>>
>>
>
>