emacs-orgmode
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Orgmode] bug: html export, latex fragments and emphasize


From: Carsten Dominik
Subject: Re: [Orgmode] bug: html export, latex fragments and emphasize
Date: Fri, 12 Nov 2010 07:22:21 -0600


On Nov 11, 2010, at 7:28 PM, Nicolas Goaziou wrote:

Hello,

Here is a problem when a latex fragment is split across two lines and
an emphasize follows. The text won't be italicized upon exporting to
HTML.

=====
* latex-fragments bug

 Imagine we have a formula starting here $e^{i\pi} +
 1 = 0$. Now we have a problem with /emphasize/.
=====

This is because the line starts with a char with 'org-protected
property and, thus, get caught by the "Protected HTML" (org-html.el
l. 1216) part of `org-export-as-html'. In others words, the line is
inserted as-is in the output buffer, before getting any
transformation.

I'm not sure how it should be done (I don't get yet the usefulness of
this "Protected HTML" part), but that piece of code may be moved after
the `org-html-expand' call, as long as every sub-function in
`org-html-expand' has a check to prevent modifying protected stuff
(this not yet the case for `org-export-with-emphasize' and
`org-html-protect' while others seem ok).

But even in this case, every function getting called after that would
be ignored. So, for example, links would not be inserted.

Couldn't the "Protected HTML" part be removed altogether?

Hi Nicolas,

I am not entirely sure if it could be removed.  The origin of this
is to make begin_html ... end_html work to keep entire blocks unmodified. It is possible that the code after this line by now has enough checks for protectedness - but I am not sure.

For the time being, I have pushed a different fix which does fix the specific issue you show, but will still fail if the math expression in more than 2 lines :(

It might be worth while to switch the protected html part off and see what happens - but that would have to be done on a special branch with significant testing by a number of people.

Regards

- Carsten




reply via email to

[Prev in Thread] Current Thread [Next in Thread]