[R-pkg-devel] HTML5 errors in win-builder R-release check

[John Fox]

Dear Duncan and John,

I also cross-posted some of the discussion on R-SIG-Mac to 
r-package-devel when it became clear that the problem wasn't necessarily 
specific to Macs. See 
<https://stat.ethz.ch/pipermail/r-package-devel/2022q2/007993.html>, and 
also most recently on R-SIG-Mac, Brian Ripley's resolution of the 
matter, at <https://stat.ethz.ch/pipermail/r-sig-mac/2022-May/014443.html>.

Best,
  John

On 2022-05-03 2:50 p.m., Duncan Murdoch wrote:
> On 03/05/2022 1:54 p.m., J C Nash wrote:
>> I've been asked to "fix" some NOTEs in two of my packages.
>>
>> The local (Linux Mint) R CMD check --as-cran gives no errors. Nor does 
>> win-builder for R-devel,
>> but R-release gives many errors of the form
>>
>> Found the following problems:
>> Rcgmin.Rd:17:1: Warning: <table> attribute "width" not allowed for HTML5
>>
>> Rcgmin.Rd:46:1: Warning: <tr> attribute "valign" not allowed for HTML5
>>
>> Rcgmin.Rd:50:1: Warning: <tr> attribute "valign" not allowed for HTML5
>>
>> Rcgmin.Rd:56:1: Warning: <tr> attribute "valign" not allowed for HTML5
>>
>>
>> In some cases the lines pointed to are beyond the end of my Rd file.
>>
>> Does anyone know the source, and hopefully the solution, to these?
>>
>> It's really difficult to fix something that does not appear to be an 
>> error
>> in the systems I can dig into.
> 
> This has been discussed on R-sig-mac.
> 
> The messages likely originate in the "tidy" utility.  Some systems have 
> very old versions of tidy which complain about very valid things; others 
> have very new versions which spot errors in the HTML generation that R 
> uses when building HTML help pages.
> 
> I think the error locations are misleading, in that "Rcgmin.Rd:17:1" 
> really should be "Rcgmin.html:17:1", i.e. column 1 of line 17 of the 
> html version of the Rcgmin help page (and I'd guess the column number is 
> wrong too).
> 
> There may be a few real errors mixed in with the spurious ones, but I 
> think I'd ignore all such warnings for now, since there's nothing you 
> can do about most of them.
> 
> Duncan Murdoch
> 
> ______________________________________________
> R-package-devel using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-package-devel
-- 
John Fox, Professor Emeritus
McMaster University
Hamilton, Ontario, Canada
web: https://socialsciences.mcmaster.ca/jfox/