I would expect this to be problematic, and I wouldn't rely on any 
predictable behaviour. This thread on a W3C list:


suggests that the idea of a text node in HTML consisting only of a 
combining diacritical mark is itself controversial.

I've been digging around in the W3C specs a little, and I can't find 
anything specifically relating to this, though -- does anyone know 
whether there is any definitive statement about the acceptability or 
otherwise of a text node which consists only of a combining diacritic 
(or begins with one)?


On 13-11-26 12:26 AM, Peter Boot wrote:
> Hello list,
>>From the department of curiosities:
> In one of the texts we’re editing, acute accents in red have been added to a base letter in black. I’m not sure whether this should work, but when I try to encode this as:
>        e<seg style="color:red">&#x0301;</seg>
> (which uses the combining character acute accent)
> Sebastian’s stylesheets create
>       e<span style="color:red">́</span>
> and I was pleasantly surprised that the major browsers understand that in spite of the intervening markup, the combining accent belongs to the preceding character. Chrome even displays the character and accent in the correct colours! (IE and Firefox don’t).
> Would you think this behaviour (merging of character and combining accent in spite of intervening markup) is something that we can rely on? Would you expect browsers to develop better support for characters and accents in different colours?
> Peter

Martin Holmes
University of Victoria Humanities Computing and Media Centre
([log in to unmask])