XPointer WD comments
Lars Marius Garshol
larsga at ifi.uio.no
Tue Aug 25 15:50:26 BST 1998
I've just returned to my XPointer implementation and updated it,
adding support for more of the specification. Because of this I have
some new questions and reflections on the draft:
- The definition of the term 'resource' in section 1.3 seems to me
both circular with 'locator' and also rather unclear.
- The WD says nothing whatever about what kinds of errors exist and
what to do with them. To me it seems sensible to operate with three
kinds of errors:
- syntactic ones, where the locator fails to conform to the
- semantic ones, where the locator does not make sense, such as
- simple failures, where the locator attempts to locate
non-existent nodes, such as child(4) on an element with three
- The WD does not say anything about how to interpret locator terms
using nodes located with the 'string', 'attr' or 'span' keywords as
the source of a locator. Ie: what is the child(1) of a span?
- Similarly, nothing is said about how to continue locating from a
list of candidate nodes, such as in root().child(all).child(1)
- According to the WD, a 'string' location term can locate a string
that stretches across several different nodes. This seems to me
both difficult to implement and somewhat unecessary, since as far
as I can see the same could be achieved by use of the 'span' term.
- Maybe a non-normative appendix to the spec should add something
about how to represent 'span' results in terms of the DOM? Should
one use a nodelist or a new node class?
- Does the A.2 section indicate that a future version of the XPointer
spec will include a formal description of XPointer semantics in
terms of the DOM?
- Nothing whatever is said about the data model on which XPointers
operate, which means that issues like entity handling and attribute
defaulting are not covered. For instance, would the XPointer
string("lars") match anywhere in the following instance?
<!DOCTYPE ex [
<!ENTITY a "a">
xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)
More information about the Xml-dev