The Unicode Consortium Discussion Forum

Does anyone know whether it's acceptable to have an XML text node which consists only of a combining diacritic? If so, what should the behaviour of a user-agent be?

For instance, in XHTML:



Would you expect the combining acute accent u+0301 to be rendered on the final e from the preceding span?

I can't find any information in XML or HTML specifications on what to expect in such circumstances.

"A combining character should be placed immediately after its "base" character, with no intervening markup or space, just as is the case for combining accents. "

From: http://www.w3.org/TR/2009/WD-xml-entity-names-20091117/

My quick take is that I would not expect your example to be robust, because of limitations in implementations, if not in the spec. By breaking up the text into <span> elements most (if not all) implementations would interrupt the text run at that point, and then the renderer cannot render the combining sequence properly - few if any implementations will have the tools to handle sequences across text run boundaries.

Something like that.