[Rd] In Rd: \describe \itemize and \item

[Duncan Murdoch]

On 1/30/2009 2:22 PM, Nicholas L Crookston wrote:
> Duncan Murdoch <murdoch at stats.uwo.ca> wrote on 01/29/2009 07:14:11 PM:
> 
>> On 29/01/2009 5:22 PM, Nicholas L Crookston wrote:
>> > I rebuilt a test version of my yaImpute package using Uwe Ligges' 
> windows
>> > build service.
>> > I got this reply: Check result: OK
>> > R version 2.9.0 Under development (unstable) (2009-01-27 r47756)
>> >
>> > Here is the issue: I have the following construct in help files (this
>> > example relates to "data"):
>> >
>> > \format{
>> >   A data frame with 847 rows and 29 columns: \cr\cr
>> >   Ground based measurements of trees (Y-variables):
>> >   \itemize{
>> >      \item{TopHt}{Height of tallest trees (ft)}
>> >      \item{LnVolL}{Log of the volume \eqn{(ft^3/acre)} of western 
> larch}
>> >  ...... many lines deleted ......
>> >      \item{LnVolPP}{Log of the volume \eqn{(ft^3/acre)} of ponderosa 
> pine}
>> >      \item{CCover}{Canopy cover (percent)}
>> >   }
>> >   Geographic Location, Slope, and Aspect (X-variables):
>> >   \itemize{
>> >     \item{utmx}{UTM easting at plot center}
>> >     \item{utmy}{UTM northing at plot center}
>> > ...... many lines deleted ......
>> >     \item{elevm}{Mean elevation (ft) above sea level over plot}
>> >     \item{slpsinaspm}{Mean of slope (proportion) times the sine of 
> aspect}
>> >   }
>> > ...... many lines deleted ......
>> > }
>> >
>> > What I get in the help file is:
>> > =========================
>> > Format  <this word is set in bold, correctly>
>> > A data frame with 847 rows and 29 columns:
>> >
>> > Ground based measurements of trees (Y-variables):
>> >
>> >    TopHt
>> >       Height of tallest trees (ft)
>> >    LnVolL
>> >       Log of the volume (ft^3/acre) of western larch
>> > ========================= ......and so on........
>> > [I've inserted 3 spaces to mimic tabs in this illustration]
>> >
>> > I would expect (and I used to get) something like this:
>> > =========================
>> > Ground based measurements of trees (Y-variables):
>> >
>> >  <bullet here>  TopHt   Height of tallest trees (ft)
>> >  <bullet here>  LnVolL  Log of the volume (ft^3/acre) of western larch
>> > =========================
>> >
> 
>> I think we need more info to comment on.  Is this file one of the ones
>> in the yaImpute package on CRAN, or is it unreleased?  Is the Rd file
>> available for us to look at?  Does the man page look different in 2.8.1
>> from what you expect (there haven't been serious changes there), or only
>> in R-devel (2.9.0)?  2.6.2 is pretty old, but I don't think any of the
>> requested changes should be incompatible with it.
> 
>> Duncan Murdoch
> 
> The problem is in an unreleased version built in the development 
> environment.
> The entire unreleased package can be found at this temporary location: 
> http://forest.moscowfsl.wsu.edu/gems/yaImpute_1.0-8.tar.gz
> 
> The original is on CRAN (version 1.0-7). It worked under the old parser, 
> but
> not using parser 2 (the warnings on CRAN point to the problem). 
> You can see the difference by looking at TallyLake.Rd 
> in the two versions. I would be happy to send you just these files if 
> you'd 
> like me to.
> 
> Note that text version of the help file builds as I expect it to. 
> The problem is in the htlm version.
> 
> PS: I tried two versions of \items 
>     \item{utmx}{UTM easting at plot center}
> and 
>     \item{utmx   }{UTM easting at plot center}
> (note added spaces, added for readability of the Rd file 
> this difference does not seem to matter.)


Okay, I think I see now.  I installed your package version 1.0-8 in 
R-devel (to be 2.9.0), and the ?TallyLake man page looks fine to me in 
text mode, looks slightly funny in HTML (with the labels one line above 
the thing they're describing).  But to me it looks better in the new 
HTML than version 1.0-7 looked in 2.8.1:  in that version, there was no 
attempt to line up the descriptions to a common left margin, so they 
tended to be confused with the thing they were describing.

So I'm not sure what you'd really like to see in the output.  Can you 
describe it again?

Duncan Murdoch