Re: [xml] ignorableWhitespace SAX callback

[Daniel Veillard <veillard redhat com>]

On Tue, Jun 29, 2004 at 09:22:47AM +0100, Chris Burdess wrote:
> Hello,
>
> Given a document without a DTD, libxml2 reports whitespace characters
> via the ignorableWhitespace SAX callback. However, as Daniel notes in
>
>   http://mail.gnome.org/archives/xml/2004-April/msg00244.html
>
> if there is no DTD, you cannot determine that any whitespace is
> ignorable. I have discussed this issue with Elliotte Rusty Harold, who

  Well there is the xml:space information too.

> is of the opinion that in such cases all whitespace should be reported
> via the characters callback. The SAX documentation is not very explicit
> on this issue. What is the libxml2 position?

  I think Eliotte is right, but it's like an extremist position, libxml2
fallbacks to an heuristic in that case, but I think in most cases
it will use the character() callback. See areBlanks() in parser.c .
IMHO legal blanks before and after the document root content should be
reported as ignorable, at least those contradict Eliotte position.

> If this has been discussed before on the mailing list please excuse me -
> the search function is 404 atm so I've relied on Google.

  there is a search on xmlsoft.org which does work and do index the
mail archives.

Daniel

-- 
Daniel Veillard      | Red Hat Desktop team http://redhat.com/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/