Re: [Docutils-users] rst2html for partial HTML generation [Was: Re: 2015: reStructuredText implementations in other programming languages?]

[Guenter M. <mi...@us...>]

Ahoj Matěj,

On 2015-06-26, Matěj Cepl wrote:

> However, I have decided to make The Right Thing™ and wrote my 
> hexo renderer using rst2html 
> (https://gitlab.com/mcepl/hexo-renderer-restructuredtext) except 
> now I have a different problem (issue #1 in that repository).  
> rst2html generates only complete HTML documents and relies on 
> tons of CSS stylesheets.

To get parts of the HTML document, the "publish_parts()" convenience
function can be used. It is described in
http://docutils.sourceforge.net/docs/api/publisher.html#publisher-convenience-functions
with example code in
http://docutils.sourceforge.net/docutils/examples.py

> Would it be possible to somehow persuade rst2html to produce 
> just as plain HTML as possible without dependency on too much of 
> CSS?

This depends on what you intend:

* the default html4css2 writer uses one stylesheet¹, default.css 
  (which is very long)
  
* the new html_base writer (!! to be renamed !!)² in the SVN repository
  uses two stylesheets¹:
  
  * minimal.css   required basic rules
  * plain.css	  rules for easier reading and pleasant layout
  
  This writer also gets rid of IE 6 compatibility hacks to achieve a cleaner
  HTML output.

* the html4trans__ writer in the sandbox does not depend on CSS
  It uses hard-coded HTML-styling (bad!) instead for rST elements that
  have no real equivalent in HTML.
  
  __ http://docutils.sourceforge.net/sandbox/html4trans/

¹ + ``math.css``, if you have maths in the document and the "math-output"
  setting is html

² cf. http://repo.or.cz/w/docutils.git/blob_plain/HEAD:/docutils/docs/user/html.txt


Günter