Re: [O] Conditional link export?

[Nicolas Goaziou]

Oleh Krehel <address@hidden> writes:

> Nicolas Goaziou <address@hidden> writes:
>
>>     The current time is: src_emacs-lisp[:results html]{(format "<a
>> href=%S>%s</a>" "http://google.com"; (format-time-string "%H:%M"))}.
>
> This is quite restrictive, since it implies that I want to export to
> HTML. What I really want is to insert a raw string, with no further
> processing into the export, whatever format the export may have.

Then use [:results raw]. This is also in the manual.

> I think it's better to have people use !(foo) than to be turned away by
> src_emacs-lisp[:results html]{(foo)} and use nothing at all.

You don't have to write [:results html] or [:results raw] for every
inline block you write. You can set the results per subtree, per
document, or globally. So really, it boils down to:

  src_emacs-lisp{(foo)}

You can even define a macro for that

  #+MACRO: eval src_emacs-lisp[:results raw]{($1)}

  {{{eval(foo)}}}

even less characters.

> Besides, Elisp is a very strong library for Org. We could also have
> e.g.
>
>     #+INLINE_LANG Elisp
>
> to set the language for !() in the current document.

Again, having duplicate syntax is out of question, IMO. I don't think
the current one is hideous either. With all due respect, the fact that
you don't like it is really insufficient as a justification for
introducing a special alternate syntax.

>> Do you want to provide a patch for it?
>
> I could provide a patch for `org-edit-src-code' to work with !().  I'm
> not interested in using the current inline syntax.

Too bad, then.


Regards,