Re: XML Schema and DTD for CVS Info in XML files?

[Herr Christian Wolfgang Hujer]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello Deron, dear list members,


Am Dienstag, 11. März 2003 23:48 schrieb Deron Meranda:
> > Currently, we use the following approach to use CVS information in XML
> > files: ...
> > <!--
> >   - $Log$
> >   -->
>
> Comments also don't work well because the "--" (hyphen-hyphen)
> sequence is illegal in XML comments.  In fact, even with elements you
> probably need to use CDATA escapes.  I would definitely place a $Log$
> inside an element body, as in
>
>    <cvs:log><![CDATA[
>    $Log$
>    ]]></cvs:log>

That's a good idea but not always possible, for instance, when there's an easy 
approach to add attributes but adding elements is very difficult.
We simply forbid the "--" in CVS Logs, as we forbid < and & and TODO.
But that's a solution working fine on a small group, whereas a widespread 
solution must of course take into account that some won't be aware of the 
"--" problem (why should they care about "--", anyway).

Thank you for the hint.


> Other tags could probably either be inside attributes or elements,
> depending upon how structured the value is.
> I do agree that XML namespaces should be used.
Yeah, I like them :-)

> One big problem with automatically inserting text into a document is
> that of character sets.  What if the XML document uses UTF-16 rather
> than UTF-8 or even ISO-8859-15?  But I don't know if that is easily
> solved.  Of course XML basically dictates that UTF-8 is the default
> unless its either explicitly specified or it can be automatically
> detected to be something else.
Thank you very much for pointing this out.
I work in an ideal environment: A Linux system completely configured to use 
UTF-8 (yeah, I do not use GNOME for that reason).
All files - Java, Perl, PHP, XML, /etc/passwd etc. are encoded in UTF-8.

Currently, if the CVS log is included in an XML document, the log's author 
must always be aware of the restriction XML might imply on the log comment.

But this problem is generic to automatically inserting bytes into a file 
format like XML, it is not specific to the combination CVS / XML, so I do not 
think it is a problem I could solve - or do you have a hint? Actually I'd not 
try to solve that problem because I see no chance to do so.


I could imagine a CVS-XML-solution with a Schema defining the simple types 
(based on regular expressions) required for the CVS Keywords, a Schema to 
choose to use attributes only, a Schema to use elements only and a Schema to 
use both - and of course a corresponding set of modules for extending old 
DTD-based XHTML / SVG.

Are you (Deron and other list members) interested?


Bye
- -- 
ITCQIS GmbH
Christian Wolfgang Hujer
Geschäftsführender Gesellschafter
Telefon: +49  (0)89  27 37 04 37
Telefax: +49  (0)89  27 37 04 39
E-Mail: address@hidden
WWW: http://www.itcqis.com/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE+bncQzu6h7O/MKZkRAiuOAKDPDUAaZm9oZLfjhcMK1jYqNVxaQQCfUcwx
AYVMvee+/rAshFF8Areb3Fg=
=xYaO
-----END PGP SIGNATURE-----