On Thursday, 27 July 2017 09:47:50 UTC+2, jfbu wrote:
>
> Le 26/07/2017 à 20:09, Christoph Buchner a écrit : 
> > 
> > 
> > On Wednesday, 26 July 2017 09:30:24 UTC+2, jfbu wrote: 
> >> 
> >> ... 
> >>     ! Undefined control sequence. 
> >>     \FNH@spewnotes ...@gobbletwo \FNH@H@@footnotetext 
> >> ... 
> >>     ! Undefined control sequence. 
> >>     \capstart ->\ifcapstart \H@refstepcounter 
> >>                                                 \@captype 
> \hyper@makecurrent \@cap... 
> >>     l.10520 \capstart 
> >> 
> >>     ! Undefined control sequence. 
> >>     \capstart ...counter \@captype \hyper@makecurrent 
> >>                                                         \@captype 
> \global \let \hc... 
> >>     l.10520 \capstart 
> >> 
> >> ... 
> >>     ! Package xcolor Error: Undefined color `OldLace'. 
> >> 
> >> 
> >>     and then we reach 
> >> 
> >>     [829] [830] [831] [832] [833] [834] [835] [836] [837] [838] [839] 
> [840] 
> >> 
> >>     ! LaTeX Error: Environment theindex undefined. 
> >> 
> > 
> > 
> > Dear Jean-Francois, 
> > 
> > thank you very much for the very thorough investigation! The confusing 
> nature of the error output remind me of my past (thesis) day when I was 
> fighting with Latex. :D 
> > 
>
> Dear Christoph 
>
> I looked a bit more and it appears the Beamer class loads the 
> hyperref package in such a context that hyperref loading is only 
> partial. As a result, a number of hyperref macros are left 
> undefined. This explains most of the errors above, including 
> the first one with  \FNH@H@@footnotetext, because that macro 
> is made an alias to the undefined \H@@footnotetext. But the 
> package footnotehyper-sphinx is fooled by the fact that the 
> Boolean \ifHy@hyperfootnotes is actually defined with value 
> \iftrue, despite hyper-footnotes related macros not being 
> defined as expected from package hyperref. 
>
> The error with undefined color is due to the fact that 
> xcolor is already loaded by Beamer class and the subsequent 
> loading by Sphinx is ignored (and the conf.py was configured 
> to pass some option to xcolor so that it would know OldLace). 
>
> Similarly hyperref options declared by Sphinx would be ignored 
> because they use \PassOptionsToPackage, but hyperref has 
> already been loaded by Beamer so it is too late. 
>
> (and it has been loaded in strange way, which causes hyperref 
> to say in the log "it was stopped early") 
>
> All of these issues could be solved by a determined LaTeX user 
> having the time. 
>
> There remains the one of missing environment "theindex". Here 
> also it can be solved by defining it oneself. 
>
> So with some work one can for example compile the Sphinx own 
> docs as a Beamer presentation (of about 900 slides...) 
>
> This will have the design of a Beamer presentation but 
> the source is lacking all of the specific Beamer commands. 
>
> It would make more sense possibly then to use rather 
> \usepackage{beamerarticle} in the preamble (I have forgotten 
> what happened when I tried[1], some of the issues above 
> were still there) rather than setting the document class 
> to beamer. 
>
> [1] ah yes, there was also an error due to option clash for 
> package color if I remember correctly 
>
> Anyway, if you really want to make a Beamer presentation 
> you need a Beamer writer on Sphinx side, not a LaTeX writer. 
>
> Because you want to be able to use all commands specific 
> to a Beamer presentation. 
>
> But the Sphinx LaTeX style file is tested and done for the 
> standard classes, as is the case of all the packages it 
> uses. It is rather surprising actually that apart from 
> the titlesec incompatibility (for which I indicated a workaround) 
> and the hyperref peculiarities and the issues with package 
> options (xcolor for example), it turns out to be almost possible 
> to use the Sphinx LaTeX writer output directly with beamer class. 
>
> > I think for the moment I won't try to convince Sphinx of my intentions, 
> and will try to make either https://github.com/myint/rst2beamer work ("A 
> docutils script converting reStructuredText into Beamer-flavoured LaTeX."), 
> or try to find willing maintainers to fix bugs in 
> https://github.com/nyergler/hieroglyph/ (it is "recommended"/mentioned in 
> the sphinx faq, after all). 
> > 
>
> Certainly an "rst2beamer" is better but I don't know how 
> that can be made to fit with auto-documenting Python code 
> for example; how could one use say autodoc: where will 
> it pause? 
> how will it split Python code documentation into frames ? 
> Or, would the docstrings contain the Beamer specific 
> syntax ? 
>
 
I don't know about autodoc, I'm sure not all features are appropriate for a 
presentation.
Other markup-to-presentation tools typically handle slide separations based 
on the section structure of the document (e.g. top-level is presentation 
title, next level presentation sections, next level slides), with the 
possibility to also use a REst transition (4+ punctuation marks) as a slide 
separator, with an empty slide title, or the title of the previous slide.

-- 
You received this message because you are subscribed to the Google Groups 
"sphinx-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sphinx-users+unsubscr...@googlegroups.com.
To post to this group, send email to sphinx-users@googlegroups.com.
Visit this group at https://groups.google.com/group/sphinx-users.
For more options, visit https://groups.google.com/d/optout.

Reply via email to