Does AsciiDoc have support for scientific papers?

[mcp]

I'm looking into using something other than LaTex directly for writing scientific papers. I'm considering org-mode and using their export to latex to pdf feature. Does AsciiDoc have anything to compare?

MCP

[Marco Ciampa]

https://asciidoc.org/latex-filter.html

--

Saluton,
Marco Ciampa

[Stéphane Gourichon]

Hi,

I've been using https://github.com/asciidoctor/asciidoctor-latex and it works (on *ubuntu 20.04), just had the following observations :

• it is based on asciidoctor not the traditional asciidoc
  
• inconsistencies between HTML rendering and PDF rendering when escaping underscores in math context (one needs one backslash, the other needs two, both reject the other option, not sure what it should be anyway),
• some options in tables that normally work are ignored, like left/right alignment
• the document style for PDF rendering is different from usual (based on AMS style), so I could not use my regular style (made for and used with the traditional asciidoc implementation)

This is only what I stumbled upon. I should probably make a short test case for bullets 2 and 3 and report to authors.

Hope this helps.

-- Stéphane

[Lex Trotman]

This ML is currently for the original Asciidoc python, Asciidoctor
discuss is https://discuss.asciidoctor.org/.

Cheers
Lex

> --
> You received this message because you are subscribed to the Google Groups "asciidoc" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to asciidoc+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/asciidoc/ae32b2e5-6d77-04fa-1013-074db12de630%40gmail.com.

[Stéphane Gourichon]

Thanks Lex for clarifying.

I'd better make clear that my whole workflow is still based on
traditional Asciidoc python implementation.

I hope mentioning the "other" and comparing here remains compatible with
a constructive conversation.

Please anyone correct me if I'm wrong.


I have tried https://asciidoc.org/latex-filter.html and this raises some
questions on the topic of "asciidoc, the language to describe a
document". These are important questions for the durability of asciidoc
the language.


On the one hand, https://asciidoc.org/latex-filter.html is a filter to
include "pictures" (that happen to be output of a LaTeX renderer) to a
regular Asciidoc document.

Benefits : "plays nice with" (does not break) existing Asciidoc language
logic or markup.

Drawback : outside those "pictures", the document remains pure Asciidoc
not LaTeX. For example, subscript/superscript have to be written
asciidoc style in text. No backslash macro at all. Greek letter? Try
copy-pasting Unicode maybe. Text with inline equation terms are probably
excluded (this is somehow frowned upon anyway in the LaTeX world). Also,
"normal" text inside LaTeX context has different appearance from normal
Asciidoc text. Numbered references to equations have to be done
Asciidoc-style also. Then what happens to the list of references at the
end of your paper? AFAIK Asciidoc (the language) does not deal with that.

My feeling is you get a two-worlds document. Parts are plain asciidoc,
in a "simple markup that can express Docbook" spirit, and they include
"pictures" that look like they are somewhat isolated LaTeX fragments.

Having written scientific papers in pure LaTeX before, I'm not sure this
is enough to write a scientific paper.

I would say latex-filter is good to include a few equations in a
primarily Asciidoc document.


(By comparison, asciidoctor-latex looks like it allows you to edit an
actual LaTeX document expressed as "mostly" Asciidoc syntax with the
addition of LaTeX $math in dollar$ stuff, \[ formulas \], LaTeX macros
and environments. Not sure I even prefer that option. You probably can
cite papers, cross-reference anywhere, because it is a LaTeX document.
It feels like a hack, breaking the Asciidoc language consistency, still
it looks like it is closer to something that could be used for a
scientific paper. You might prefer it, especially it you master LaTeX
already. Be prepared to rework your LaTeX preamble if needed, and do not
bet on being able to easily recompile your nice paper from source ten
years from now. The fact that escaping underscore character has to be
done differently depending on whether you render to PDF of to HTML feels
like a telltale warning sign.)


As for me, what I need is to avoid reworking my Asciidoc-python-based
processing chain (working since ca. 2012). I'm mostly writing quotes,
invoices, technical documentation in Asciidoc. Ability to insert some
equations here and there is a nice addition, so
https://asciidoc.org/latex-filter.html is the way to go for me.


-- Stéphane

Software freelancer in Paris, maybe you need my services?
https://fidergo.fr/

[Derek Jones]

Stéphane,

> I'd better make clear that my whole workflow is still based on traditional Asciidoc python implementation.

Scripts can be integrated into the Asciidoc workflow:
https://shape-of-code.coding-guidelines.com/2019/08/11/my-books-pdf-generation-workflow/

to produce all manner of layouts, but some fiddling is needed to
add numbering to equations:
http://www.knosof.co.uk/ESEUR/ESEUR-draft.pdf