[
Lists Home |
Date Index |
Thread Index
]
Uche Ogbuji wrote:
> I know most people on this list know this, but because the above is a
> good reference post, I wanted to clarify that using numerical characters
> references does not solve #3.
>
I know. My point was meant to be, though perhaps it could have been said
better, that if you're iterating through a chunk of text and inspecting
each character, it's easy to spot the ones that are completely illegal
and throw an exception or take other corrective action. However, if you
just take a string and wrap it in <![CDATA[ and ]]>, then you'll miss
such possible illegal characters.
Depending on the language there may also be a point 4: mismatched
surrogate characters and non-characters. These are allowed in Java
strings, but not in XML data.
--
Elliotte Rusty Harold elharo@metalab.unc.edu
XML in a Nutshell 3rd Edition Just Published!
http://www.cafeconleche.org/books/xian3/
http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim
|