Re: [css-selectors] Selecting non-element nodes

Le 26/07/2013 17:20, François REMY a écrit :
>> I’m not super convinced. It seems that these examples would be
>> used between "block-level" siblings, where you rarely have "naked"
>> text nodes. (Often, there is at least a  element.)
>
> Given I just spent some time writing a function that walks through
> the DOM to find loose text nodes which I would like to hide but
> cannot remove from the DOM nor target via CSS (ie: save their
> nodeValue in an expando and basically set their nodeValue to an empty
> string, or vice-versa),

Can you explain why you want to hide text but not remove it from the DOM?


> I believe in my use case sufficiently.

Great! Please don’t forget to tell why to the people who aren’t in your 
head :)


> It's not the first time this issue comes on the table,

Do you have a pointer? I believe that the use cases would be very 
different in stylesheets vs. APIs, maybe the solutions should be as well.


> but it seems
> nobody ever proposed a solution to it, which is I believe the only
> way to hope for a fix to land at some point.
>
> However, I agree that in the read-more case you usually have <p>
> wrappers, at least in the case of a blog.

Ok. I find the use case story minor compared to the proposal actually 
making sense. I think it would help if you have a more concrete proposal:

Exactly what types of nodes would this apply to?
How do you define a node? Eg. is this more than one node? Foo<!-- -->bar
What are the semantics when used in a stylesheet? In selector APIs?

-- 
Simon Sapin

Received on Friday, 26 July 2013 16:33:29 UTC