"NAKAMURA, Hiroshi" <nahi / keynauts.com> wrote in
<JIEJKAMAONAMHNNAOOOLOELDCFAA.nahi / keynauts.com>: 

>Sorry for RWiki specific topic...

and long ;-)

>
<<SNIP>>
>> <rbi_code: >>ruby code goes here<< >
>

OK, maybe my example was missleading (because of the << and >>):

<rbi_code: 
class a 
  def say()
    p "hello"
  end
end
>

This looks like your img-example and the visitor could be written as
proposed in my previous post. 

So if you decide to add <:>-tags to rwiki we could easily add
<rbi_code:>! As I mentioned in the post before, I have concerns in what
happens if the code includes a ">". Will the RD parser fail? 

><...> is a (X)HTML and not a part of RD...
>RD has a "reference" syntax like ((<...>)) but its
>semantics is a reference.  rbi_code is for program code
>and should not be a reference I think.

I agree

>
>In RD, syntax (({...})) is for program code.
>http://www.jin.gr.jp/~nahi/RWiki/?cmd=view&name=RD+format
>How do you feel below syntax?
>
<<SNIP>>
>
>You can override apply_to_Code in rd2rwiki-lib.rb to get
>the program code ('print ">"' in above example) and insert
><textarea /> instead of '<code>print "&gt;"</code>'.

should we really override this RD-semantic? 
How about adding a new RD-syntax element, i.e. (([])) or ((~~))? 

<<SNIP>>
>
>Next candidate.  How do you feel this?
>
>  = this is a sample code
>  Using verbatim block.
>    rbi_code:
>    class Foo
>      def say
>        "hello"
>      end
>    end
>    Foo.new.say
>
>This is exactly what [ruby-talk:11208] does.
>RWiki does not support this type of extension now...
>

looks fine, too

>And.
>
>Clemens, is (X)HTML style <rbi_code>...</rbi_code> a
>must?  

absolutely not

>If you wanted dynamic program code evaluation per users
>request, we must hooked step (3).
>

true

Nahi, you tell me when you think it is time to integrate. Having
"verbatim blocks" in rwiki (RD) might be such a milestone (or the
additonal RD syntax element). Clemens