I see, so like if an element has element-only content, then we would insert 
spaces, but if it has mixed- or text-only content (simple content) the we would 
not touch it. Would this be based on the Schema or on the actual content of the 
element, you think? The reason I am asking is because you can create and 
manipulate invalid documents via XMLBeans. We would also need to find a way to 
cleanly expose this option without impacting the current functionality.
Gretchen, it would be useful to have a JIRA improvement request for this, with 
the priority you think it has, then of course, we appreciate all the help we 
can get on it.

Thanks,
Radu

-----Original Message-----
From: Mark Lewis [mailto:[EMAIL PROTECTED]
Sent: Monday, July 11, 2005 4:55 PM
To: [email protected]
Subject: RE: Whitespace is not preserved


Hmmm, I wonder if it would be possible to alter the pretty printer to
work even in this case.

What if we check whether an element has leading/trailing whitespace that
needs to be preserved, and selectively disable pretty-printing for that
element only?  Whitespace between tags wouldn't need to be preserved
(unless it was defined in the schema, but most aren't written that way),
so the overall document could still look pretty good.

-- Mark Lewis



On Mon, 2005-07-11 at 16:23 -0700, Radu Preotiuc-Pietro wrote:
> Well, to be honest, the meaning of pretty print is that you give up
> space preservation in order to get a document that's easier to read.
> If you care about spaces being preserved, you should _not_ use pretty
> print. They're exclusive and I am not sure how would they work
> together.
>  
> However, if you access the document using, say, XmlCursor, then all
> the original spaces in the infoset will be there, regardless of the
> Pretty Print value.
>  
> Radu



---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to