Re: [xml-dev] parser error : Premature end of data

[Thufir <hawat.thufir@gmail.com>]

On Sun, Nov 7, 2010 at 4:50 AM, G. Ken Holman
<gkholman@cranesoftwrights.com> wrote:
> At 2010-11-06 21:23 -0700, Thufir wrote:
[...]

> If you start two elements, you have to end two elements.  But in doing so
> you are changing the structure of your information.

Ok, that explains it; thanks.  I was kinda curious about the two open
book elements, should've tried closing another book element.

> I understand wanting to break the start tag into two lines to line up the
> attributes (Saxon does a nice job of this in its serialization).  Why do you
> want two elements?  That would break any document-element-based XPath
> expressions written for an instance with only one <book> element.
[...]

By two elements, this means a node of books within a node of books,
which you're saying will break XPath and, presumably, other processing
of the document?  I can see that it would be difficult to parse if
it's unanticipated, which I hadn't considered.  Would it make sense to
use something like:

<?xml version="1.0"?>
<library xmlns="urn:loc.gov:books">
  <book xmlns:isbn="urn:ISBN:0-395-36341-6">
    <!--- omitted  -->
  </book>
</library>


so that the library and book namespace vocabularies (?) are available
to the entire document?  Or, is it better to have the default and
other broad-scope vocabularies encapsulated within a single node?



thanks,

Thufir