Lists Home |
Date Index |
On Fri, Mar 21, 2003 at 08:15:24AM -0500, Elliotte Rusty Harold wrote:
> Keep in mind that it is possible *now*, *today*, in XML 1.0 for a
> file or a stream to contain multiple consecutive XML documents, as
> well as other things. In fact, this may actually occur with multipart
> MIME and/or HTTP 1.1. XML 1.0 makes no requirement that a file or
> stream contain only one document.
> An XML parser is typically positioned at the byte that begins the XML
> document by some external process. There is no reason that the
> initial parser position has to be byte 0 of the stream. For instance,
> in HTTP 1.0 it's almost never byte 0 of the stream. An external
> process needs to strip off the header first.
The key is that the layer calling the parser must stop parsing
at the right boundary requiring some kind of preparsing or a-priori
knowledge on the stream. If the parser is left going on the stream then
it must report an error and abort.
Daniel Veillard | Red Hat Network https://rhn.redhat.com/
email@example.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/