mozdev.org

• Saturday November 21st 2009
• All the leaves are brown …

Skip to Content

XPointerLib    

resources:	Home Release Notes Installation API / Examples Source Code Bugs Links Mailing List

Caveats and Notes for Releases

Version 0.2.8

• Fixes a bug with XPointer creation.

Version 0.2.7

• Firefox 3 compatibility.

Version 0.2.6

• Add "hide node" functionality. (Never formally released, skipped to 0.2.7.)

Version 0.2.5

• Changes by C. Greg Hagerty. (Sorry, don't have more details on these changes.)

Version 0.2.4

• Firefox 2 compatibility.

Version 0.2.3

• Fixed a bug where unknown schemes caused errors (they should be skipped). Bug reported and fixed by Eliot Setzer.

Version 0.2.1

• Fixed a small bug in XPointer creation.

Version 0.2

• Added markElement method, which allows a client to mark an element before it is inserted into the DOM. Elements so marked will be ignored by subsequent XPointerLib processing. This means that you can insert multiple icons, etc at the locations resolved by XPointerLib, and each new insertion will behave as though the document were still in pristine form.
• Fixed a bug with createXPointerFromSelection and createXPointerFromRange where the returned xpointer was not properly wrapped in xpointer(...)
• Added getVersion method to get the version of XPointerLib that you are using

---

Version 0.1.4

• Fixed bug 2857, enabling createXPointerFromSelection to work for empty selections.

Version 0.1.3

• Fixed differentiation between XML and HTML Documents (it should work now). We now test for the existence of the document.body property, which only exists in HTML DOM Documents. HTML DOM Documents get lowercase'd comparisons, XML DOM documents do not.

Version 0.1.2

• Added differentiation between XML and HTML documents in order to deal with their opposing stipulations about case sensitivity for tag names. For HTML documents, tag names are now handled in a non-case-sensitive manner. For XML documents, tag names are handled in a case-sensitive manner. Note that XHTML is an XML-derived language, and is processed as such.

Version 0.1.1

• Fixed bug 2679, enabling createXPointerFromSelection to work on backwards-selected text

Version 0.1

• This version is noticably slow when used with a large number of XPointers. For example, parsing all of the XPointer annotations stored on the W3C's annotations server that pertain to www.w3.org/ (some 65-70 annotations as of this writing) will cause the UI to become unresponsive for about 4-5 seconds on a Pentium III / Linux box. I intend to do some profiling to determine candidates for optimization, and ultimately to move to C++ (speed may be a limitation I can't escape in JavaScript).
• As mentioned on the main page, the resolver only handles a subset of the total XPointer langugage. This subset is large enough to capture any XPointer generated by the Amaya or Annozilla annotations client. It does not include most of the axisName specifiers (e.g. descendant, parent, self), though it does include the abbreviated form of the child axis specifier. It should, however, fail with a useful JavaScript error whenever someone tries to use an unimplemented piece of XPointer.
• There may be other issues as well. If you find anything, please let me know by submitting a bug. Thanks!