I have tested the XHTML5 XSD found at http://www.xmlmind.com/xhtml5_resources.shtml, in another XML editor. The editor’s validity service complained about a non-determinist content model defined at line 1157. And, at that location in the XSD, there is even a comment which notifies about the issue:
<!-- XERCES: cos-nonambig: tfoot violate "Unique Particle Attribution" --> Becuase the XSD apparently tries allow <tfoot> to occur EITHER in the start of the table, or in the end of the tale (when we look at the DOM - visually it always occurs at the end.) This is done by including both <xs:element ref="tfoot"/> right after the <thead>, as well as a <xs:element ref="tfoot" minOccurs="0"/> after last <tbody>. The issue here is that while HTML4 and XHTML5 allowed the <tfoot> element to occur at the beginning of <table> element (after <thead> but before <tbody>), HTML5 allows it to occur EITHER at the traditional location OR *after* the last <tbody>. The two variants cause different DOMs, but to the user they ought to be identical. To which degree that breaks the XSD spec, I don't know. But when I commented out the last (<xs:element ref="tfoot" minOccurs="0"/>), the schema started to work just fine. I am not sure that it is necessary that the schema allows <tfoot> to occur in both contexts (the one or the other). Per HTML5, it is semantically equivalent regardless where the tfoot occurs. -- leif halvard silli -- XMLmind XML Editor Support List [email protected] http://www.xmlmind.com/mailman/listinfo/xmleditor-support
