Hey all, I'm not sure what the standard defines, but my experience with Xerces differs from this.
> On 16 December 2001 04:23, Dean Roddey [SMTP:[EMAIL PROTECTED]] > wrote: > > Yeh, that's true. Normalization is done at the low character > > spooling level, so its done always really. The only time its not > > done at that level is when what's being spooled is an internal > > entity, which has already been normalized during its initial > > parse, so one way or another the normalization is always done. > > > > I guess I was just trying to make sure he understand that CDATA > > means two different things, since he seemed to be thinking > > that. I just picked a really bad example to try to point it out > > :-) I hadn't realized that they had overloaded the term, thanks. Seems like a bad idea doesn't it? > > ----- Original Message ----- > > From: "Hal DeVore" <[EMAIL PROTECTED]> > > To: <[EMAIL PROTECTED]> > > Sent: Saturday, December 15, 2001 6:40 PM > > Subject: Re: Reading attribute values with leading spaces > > > > > > > > > > >>>>> On Sat, 15 Dec 2001, "Dean" == Dean Roddey wrote: > > > > > > Dean> not imply the same thing that a *CDATA Section* means in > > > Dean> the body of an XML document, which means "don't be messin > > > Dean> with my stuff". > > > > > > My understanding is that whitespace normalization > > > happens even in CDATA _sections_. Not in Xerces it doesn't. If I have a CDATA attribute all whitespace is normalized, but if I have a CDATA section within an element whitespace is left alone. At least it is in SAX2. jas. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
