RE: Domain mapping implementation

> I see two places where it could be done (they are not 
> mutually exclusive):
> 1.- First, remove the possible duplicated values of the 
> input domain string. That is applicable, for example, for 
> the comma separated values in a meta keywords HTML tag.
> 2.- Remove the possible duplicated values of the resulting output 
> domain string, after applying the mappings.

So we should probably have add a step in the spec, just before the current step 4 (Return the resulting string), that instruct to remove duplicates there, as the first case should be also handled by the second one.

Sounds fine to me.

And about the quotes, the more I think about it the more it seems that having a step to trim them, even if the HTML5 spec don't mention them is probably better. I can't imagine someone wanting to have a real domain value that is bracketed by quotes.


Received on Tuesday, 30 October 2012 12:52:12 UTC