Georgi Boshnakov
2018-Oct-29 08:57 UTC
[Rd] disappearing paragraphs in html rendering of text from Rd macros
Recent changes in R-devel seem to change the HTML rendering of text returned by
Rd macros,
in that now empty lines do not cause paragraphs to end (the empty lines are kept
but <p> tags are not set).
As a result, multiple paragraphs are output as a single one.
A complete example Rd file is at:
https://github.com/GeoBosh/reprexes/blob/master/disappearingParagraphs/a.Rd
rendered (2018-10-28) with R-devel:
https://github.com/GeoBosh/reprexes/blob/master/disappearingParagraphs/a.html
and R-3.5.1patched:
https://github.com/GeoBosh/reprexes/blob/master/disappearingParagraphs/a351.html
The exact system information used for rendering is in the above html files.
In essence, the following:
\newcommand{\abc}{\Sexpr[stage=build,results=rd]{paste0(letters[1:4], collapse =
"\n\n")}}
produces in R-devel something like
<p>a
b
c
d</p>
where the empty lines may be distorted in this email, while in R-3.5.1 the empty
lines cause paragraph tags around each chunk:
<p>a</p>
<p>b</p>
<p>c</p>
<p>d</p>
Directly embedding the Sexpr in the Rd file also produces all <p> tags in
R-devel and R-3.5.1patched.,
---
Georgi Boshnakov
Martin Maechler
2018-Oct-30 11:06 UTC
[Rd] disappearing paragraphs in html rendering of text from Rd macros
>>>>> Georgi Boshnakov >>>>> on Mon, 29 Oct 2018 08:57:50 +0000 writes:> Recent changes in R-devel seem to change the HTML rendering of text returned by Rd macros, > in that now empty lines do not cause paragraphs to end (the empty lines are kept but <p> tags are not set). > As a result, multiple paragraphs are output as a single one.> A complete example Rd file is at: https://github.com/GeoBosh/reprexes/blob/master/disappearingParagraphs/a.Rd> rendered (2018-10-28) with R-devel: https://github.com/GeoBosh/reprexes/blob/master/disappearingParagraphs/a.html> and R-3.5.1patched: https://github.com/GeoBosh/reprexes/blob/master/disappearingParagraphs/a351.html[..................]> Georgi BoshnakovThank you. I can reproduce what you report, thanks to your 'reprex', and will have a look in order to fix this regression in R-devel. Martin Maechler