sgml.g and tree structure for nested tags

Tue Aug 10 17:31:05 UTC 2010

Ruan Strydom, 10.08.2010:
> The formatters (little buggy, and much simpler) currently works similar
> there is no concept of a tree. It was one of my original ideas to simply
> have a list, but after some uneducated thought, I figured it would
> simplify the context builder. But the builders and the .g became such a
> complex mess that I see this approach actually only simplifying it.
> 
> Sorry I do not have any time for the next 3 months, I am writing exams and
> lots of work.
> 
> End Nov and Dec I will have time again for some hobby code'n.

No worry, I'll continue to work on Quanta after GSOC I hope so I'll try to do 
that.

Cheers

> On Sunday 08 August 2010 16:46:58 Milian Wolff wrote:
> > hey Ruan,
> > 
> > I'm thinking about simplifying sgml.g by making it generate a list
> > instead of a tree. I remember that you have asked us/me back then how to
> > create the tree and put much work into it, but when I actually work with
> > it now, I don't think a tree is really that good here.
> > 
> > Html/Xml is pretty easy after all, and the parser should just give us
> > nodes with stuff like
> > 
> > - <!doctype
> > - <?xml ?>
> > - <opentag> (attributes, maybe empty)
> > - </closetag>
> > - CDATA
> > - TEXT
> > 
> > that's basically it. The rest should go into the ContextBuilder, i.d.
> > open a context inside <opentag>, close it on </closetag>.
> > 
> > When doing so, we can also report proper errors when tags are incorrectly
> > nested, or closed at the wrong point (or not at all).
> > 
> > Another point, which actually sparked my idea, is this:
> > 
> > <html>
> > <?php /* ... */ ?>
> > </html>
> > 
> > I personally would see this as HTML embedded in PHP (as the latter is the
> > preprocessor). The current multilang approach hands "<html>\n" and
> > "\n</html>" as two parts over to the XML language plugin, hence the AST
> > cannot be built in the correct way currently. When it's simply generating
> > a list of stuff and the context builder would cope with the rest as
> > proposed above, it should work like a charm.
> > 
> > Bye

-- 
Milian Wolff
mail at milianw.de
http://milianw.de
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kdevelop-devel/attachments/20100810/fa6a9bd4/attachment.sig>