RE: XPath conformance? was RE: [xml-dev] storing XML files

[Evan Lenz <elenz@xyzfind.com>]

Mike Champion wrote:
> I'd submit that a processor that executes legal XPath expressions "is" an
> XPath processor, whatever it's called, and whatever superset of
> XPath syntax
> it also supports.

Because XPath can be extended via an extended function library, a la XSLT
and XPointer, I would disagree with the claim that a processor that supports
an extended *syntax* is in compliance with the XPath specification.

There are parts of the picture that are not standardized--most notably how a
node-set (or even a boolean, for that matter) should be serialized. These
are implementation-dependent (whether the implementation is the XSLT
specification or some XML database vendor).

This leaves vendors much freedom (at the expense of having a standard).
However, they should not feel free to extending the syntax of XPath, i.e.
what the XPath spec *does* standardize.

For example, the following is not in compliance with the XPath spec:

/customer[name ~= "Mike"]

whereas this alternative expression would be perfectly fine:

/customer[ext:like(name, "Mike")]

Evan Lenz
XYZFind Corp.