This note is a byproduct of some web searches seeking out technologies in support of projects such as Open HyperScope.

The W3C definition of Xpath describes a language to locate specific parts of an html, or xhtml page when part names have not been embedded in that page by its author. This XPath tutorial is presumably relevant. This introduction warns of severe prerequisites including XHTML and XML Namespaces. XPath 2.0 is a recent new edition of the standard. There are implementations of Xpath in the XSLT language. XLink is another proposal that does not seem to support our goals.

XPath seems to be about navigating within a document. XPointer perhaps combines this with the URL to locate a part of a page on the web. The Document Object Model XPath seems designed give programers access to XPath function from several languages.

Demos

Brad Neuberg refers me to this XPath tutorial. This indeed confirms that Firefox has the requisite XLTS functionality to run this version of XPath. The missing functionality seems to be enabling XLTS code to pull the web page in which the author wishes to navigate. See plausible reasons why Firefox won’t do this.

Implementations

This and this claim to be PHP implementations of XPath. I though PHP ran on the server. A server based intermediary could solve our problems. This seems to a XSLT version. Perhaps Pathan is an open source implementation.

XPointer

As far as I can see from the abstract in the W3C XPointer documenter, XPointer is a way to help programs extract information from XML documents. It may not be strategic for our purposes.

Here is a list of annotation type projects. I kibitzed on CritSuite.