Hi Torsten,

First of all the problem of quotation marks inside
cdata does not happen to me. I tested the latest
version of DocBookWiki (from subversion)
in Fedora5. So, either you don't have the latest
version of DocBookWiki or you don't have Fedora5.

I remember that sometimes in the past I had problems
with quotation marks, so I introduced some code
to fix it (by escaping them). However it turned out
that the problem actually was either with the
version of the browser or with the version or the
configuration of apache, so the fix was actually
*creating* problems (by introducing extra slashes).
So I removed it. 

About CDATA, I think that introducing
--cdata
----
would be OK. The parser does support nested blocks,
however, CDATA blocks, being somewhat different from
the XML elements, are handled in a special way;
they have to be preprocessed before the text/wiki
is submitted to the parser (this is what is done
currently with the <![CDATA[...]]> blocks).

So, the parser will never have the chance to parse
the block --cdata, because it has been already
removed by the preprocessing stage. (There is
also a post-processing stage which puts them back
after the text/wiki is converted to XML/DocBook.)

Since the matters are like this, then we don't
have to use a block-similar markup for denoting
CDATA sections. Any markup that does not create
ambiguity would be OK. So, I would propose to
use something like this: --[[xyz]]--
hoping that '--[[' is not so common in the text
of documents and ']]--' may not be so common
in the content of cdata sections. Do you have
any better suggestion about these markups?

They don't have to be symetric, they just have
to be simple and unambiguous (otherwise, if they
are not simple, then we better use <![CDATA[...]]>
itself). I think that it is also better to be
inline markup (instead of block markup), since
<![CDATA[...]]> itself is inline.

Then, going a step further, to make things simple
(for the user, the documentation writter, not for
us the programmers), we can define implied cdata
sections to be like this:
This block:
--blocktag--
abc
xyz
----
is equivalent to this:
--blocktag
<!CDATA[abc
xyz]]>
----

I think that it can be usefull, in general.

Regards,
Dashamir


--- Torsten Schlabach <[EMAIL PROTECTED]> wrote:

> Hi Dashamir,
> 
> > About the suggestion, maybe it is better to keep
> > --code unmodified and to introduce another tag,
> > e.g. --code-- which would wrap the content in 
> > CDATA. What do you think?
> 
> I just had a look at "DocBook: The Definitive Guide"
> and you're probably right. There should not be an
> automatism to have a programlisting wrapped into a
> CDATA tag in any case because the docbook standard
> does allow for markup inside a program listing,
> which can be useful.
> 
> But then, why not keep it simple and introduce
> 
> --cdata
> ----
> 
> so it would be
> 
> --code
> --cdata
> your program listing here
> ----
> ----
> 
> That would also allow for something like:
> 
> --code
> --cdata
> some not so important code in the beginning
> ----
> <emphasis>
> --cdata
> more code
> ----
> </emphasis>
> ----
> 
> Or would nested blocks cause a problem for the
> current version of the parser?
> 
> WDYT?
> 
> Regards,
> Torsten


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Doc-book-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/doc-book-users

Reply via email to