Re: No More SGML

[Daniel Veillard <veillard redhat com>]

On Mon, Sep 22, 2003 at 11:02:19AM -0500, Shaun McCance wrote:
> On Mon, 2003-09-22 at 11:32, Dave Malcolm wrote:
> > Possibly heading off-topic, but can I ask what tools people use to do
> > this conversion?
> > 
> > Within Conglomerate I've got an SGML->XML importer.  It's a wrapper
> > around sgml2xml, but it doesn't work particularly well.
> > 
> > So I'd welcome suggestions on how to improve this...
> > 
> > If you're feeling particularly nice, please add them as comments to this
> > bug:
> > http://bugzilla.gnome.org/show_bug.cgi?id=122651
> 
> Probably the best hope of an automated SGML->XML conversion is to use
> the DocBook parser in the current libxml2.  It will parse into the same
> internal structures used for XML.  Then just save that out to file.  If
> libxml2 fails, then your best friend will be emacs.  A clever awk user
> could probably manage to script some of it.

  Well libxml2 didn't handle the minimization features correctly, and
I think external entities parsing was not working correctly either.
To me that's why the previous code was really too bad to be maintained.

> I think you're talking about having Conglomerate able to import SGML. 
> My suggestion is to give up.  The whole problem with SGML is that it's
> just bloody impossible to write parsers that work.  XML was created
> specifically to address this.

  Exactly. I suggest using James Clark tool for the conversion, it
will loose the entity structure, but it should keep the content safe
at least.

> Really, it's passed time people just stopped using SGML.  XML has been
> such a buzzword for so long, I'm surprised that people are still using
> stuff like DocBook SGML or HTML sans the X.  SGML has been a burden on
> programmers (and everyone else) for too long.

  I agree too, we should just do pure conversion of documents. Not look
into maintainance mode for any of those SGML docs at this point.

Daniel

-- 
Daniel Veillard      | Red Hat Network https://rhn.redhat.com/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/