Re: [xml-dev] xml 2.0 - so it's on the way after all?

[Toni Uusitalo <toni.uusitalo@pan.nu>]

At 02:48 4.2.2005, you wrote:

>If there were to be an XML 2.0, one of the goals should be that any 
>off-the-shelf parser generator can produce a parser for it. It's 
>ridiculously hard to write a correct XML parser; it ought to be 
>ridiculously easy.
>I also agree with whoever wrote that DTDs, validation and macro expansion 
>should be pulled out of the base specification entirely. Hey, here's an 
>idea: what you see is what you get.

Somewhat relating to this, an example of how hard it is to get DTD 
parameter entity expansion right; this is something I've encountered when 
writing my Parsifal parser:

The following document is well-formed (and valid too but that's not the 
point) but very troublesome to parse for MSXML (IE6 crashes) and Expat. 
Works ok with RXP and Parsifal - haven't test others.

You can test this with your browser in the link 
http://www.saunalahti.fi/~samiuus/toni/xmlproc/parsifal_tests/output_tests/output_PE1.xml

Or just examine this:

output_PE1.xml:
<!DOCTYPE doc SYSTEM "output_PE1.dtd"[]>
<doc/>

output_PE1.dtd:
<!ENTITY % b "test">

<!ENTITY % ws "">
<!ENTITY % e SYSTEM%ws;"output_PE1.ent">

<!ENTITY % PE "&#37;">
<!ENTITY%PE; s "%e;">

<!ENTITY u "%s;">
<!ELEMENT doc ANY>
<!ATTLIST doc att CDATA #FIXED "&u;">

output_PE1.ent:
%b;

Quite evil document? not very "real world" one I guess but well-formed 
nevertheless.


with respect,
Toni Uusitalo