> It's not a bug, it's the documented behavior:
>
> :first
> Matches the first selected element.http://docs.jquery.com/Selectors/first
I cite from the link above: Finds the first table row. $
("tr:first").css("font-style", "italic");
Seems quite clear
>
> Let's go through it again. When you call $(this).parents()
Sorry but I don't call .parents(), I call .parents('dl:not(:first)')
This is not the same
When parents is called is called with an expression that filters the
result list
In the :first page example, for example
$("tr:first") is called in the context of document
This call filters from all the elements the first tr not the first
table
If we apply your logic $("tr:first") will return the first table
because is the first element in the list of elements in the tree
The logic seems to be that parents('dl:not(:first)') first filter the
dl elements from the parents list and then filter for :not(:first)
If this is not the real behavior (and then there are a bug) jQuery
breaks a logic that I think we could ask for with a poll
I bet you that your logic don't win
>
> you get this array of elements:
>
> [0] <dd>
> [1] <dl>
> [2] <dd>
> [3] <dl>
> [4] <dd>
> [5] <dl>
>
> Which are all the element's parents, in order. That's the collection
> you're dealing with.
> Now, :first refers to the first element ([0]) in that set, and only
> that, it has no other meaning. So dl:first will match nothing, as no
> dl is the ':first' element. "dd:first" will match the same as dd:eq(0)
> or simply :first.
>
> You have to take in account that the pseudo-selectors are a set of
> defined filters, there is no 'semantic' in them. Maybe this could be
> improved, but that would be a change in syntax, not a 'bug
> correction'.
>
> Again, that's what slice() is for, slicing the array of elements as
> you wish.
>
> cheers,
> - ricardo
>
> On Feb 3, 12:28 am, Garito <gar...@gmail.com> wrote:
>
> > Sorry, Ricardo, but this isn't so much logic, isn't it?
>
> > When I see dl:not(:first) I read give me all dl except the first one
>
> > Imagine this was a mathematics expression
>
> > We read from left to right: first filter all dl's with the result
> > exclude the first one
>
> > I can understand a bug but it's dificult to me to understand your
> > explanation about this issue even when I test the code with the iPhone
> > safari (on Wednesday I could try it with IE to see the dl supposed, on
> > FF or Safari don't appears, problem you point)
>
> > I don't know who decides (at jQuery) what is a bug and what isn't but
> > in my opinion this is one of them
>
> > If jQuery people wants to correct them, ok
> > If not ok but jQuery will have it's first point to leave it as soon as
> > a better library appears
>
> > I can do anything more that change my code to solve this jQuery bug
> > but this don't change the fact that there is a bug at jQuery
>
> > Thanks
>
> > On 3 feb, 02:32, Ricardo Tomasi <ricardob...@gmail.com> wrote:
>
> > > I haven't seen your page, but I know that styling <dd> and <dt>
> > > elements for IE is a pain. And the invalid mark-up might bring you
> > > problems with different browsers. The usual behavior for invalid
> > > nesting is to close the offended tag to make it valid, so this
>
> > > <dl>
> > > <dt>
> > > <div></div>
> > > </dt>
> > > <dd/>
> > > </dl>
>
> > > would become
>
> > > <dl>
> > > <dt>
> > > </dt>
> > > </dl>
> > > <div></div>
> > > <dl>
> > > <dd/>
> > > </dl>
>
> > > Firefox is surprisingly tolerant in this case, but I haven't tested
> > > elsewhere. Valid mark-up is a safe harbour specially with regards to
> > > DOM manipulation.
>
> > > That aside, I found the issue, it's actually quite simple.
>
> > > this.parents('dl:not(:first)')
>
> > > let's see what's happening:
> > > - first, this gets you all the parents() for 'root'. Without any
> > > filtering, that would be DD, DL, DD, DL etc.
> > > - then you ask it to filter the elements which are 'DL's AND are not
> > > the first match. Here's the catch: :first refers to the first match in
> > > the set, which is a DD element, not a DL one. The DL is never going to
> > > be the first match.
>
> > > :first will always filter according to the current set, before any
> > > other expressions in the same call are evaluated. So what you need to
> > > do is filter the parents set again:
>
> > > var slots = $(this).parents('dl').filter(':not(:first)').map(function
> > > () {
> > > return $(".Texto:first", $(this)).text();
>
> > > }
>
> > > Or use slice() instead as I posted before, that will be much faster as
> > > you're just slicing the array without dealing with elements/selectors.
>
> > > var slots = $(this).parents('dl').slice(1).map(function() {
> > > return $(".Texto:first", $(this)).text();
>
> > > }
>
> > > cheers,
> > > - ricardo
>
> > > On Feb 2, 11:35 am, Garito <gar...@gmail.com> wrote:
>
> > > > Sorry Ricardo but you say the problem is the dt element?
>
> > > > With this markup I can do what I need in terms of css
>
> > > > With the css I put the divs inside the dt element to put a label, the
> > > > icon and the other div to switch between visible/non visible dd
>
> > > > How do you think I could put my markup to avoid this problem?
>
> > > > I don't like your suggest because one of the main reasons to change
> > > > from prototype to jquery was the css's selectors and they power
>
> > > > I thinks this is a jquery's bug, isn't it?
>
> > > > On 2 feb, 10:02, Ricardo Tomasi <ricardob...@gmail.com> wrote:
>
> > > > > Ah that's quite confusing mark-up. A <dt> element can only contain
> > > > > inline elements, certainly not divs and other definition lists. Nested
> > > > > definition lists make no sense! I couldn't find the problem.
>
> > > > > Try $(this).parents('dl').slice(1) or $(this).parents('dl').slice
> > > > > (0,-1), that will probably work.
>
> > > > > cheers,
> > > > > - ricardo
>
> > > > > On Feb 1, 10:57 pm,Garito<gar...@gmail.com> wrote:
>
> > > > > > Gracias, Ricardo!
> > > > > > I change the code with the complete test case for my issue
>
> > > > > > Sorry for the identation but the original code is generated and I
> > > > > > copy/
> > > > > > paste with firebug
>
> > > > > > If I'm not wrong, with the selector 'dl:not(:last)' SITE dl is
> > > > > > incorrect in the returned list because is the last dl (if not 'dl'
> > > > > > will be equal to 'dl:not(:last)')
>
> > > > > > am I confused or there is a bug?
>
> > > > > > Thanks!
>
> > > > > > On 1 feb, 21:07, Ricardo Tomasi <ricardob...@gmail.com> wrote:
>
> > > > > > >http://jquery.nodnod.net/cases/85
>
> > > > > > > Parece ok aqui. Pode ser alguma outra coisa na tua página, não dá
> > > > > > > pra
> > > > > > > saber sem ver o html.
> > > > > > > --------
> > > > > > > Seems to work fine here. Could you post a complete test page
> > > > > > > showing
> > > > > > > this issue?
>
> > > > > > > - ricardo
>
> > > > > > > On Feb 1, 1:25 pm,Garito<gar...@gmail.com> wrote:
>
> > > > > > > > Hi!
> > > > > > > > Please, consider this code:
>
> > > > > > > > $.fn.url2 = function(absoluta) {
> > > > > > > > var slots =
> > > > > > > > $(this).parents('dl:not(:last)').map(function() {
> > > > > > > > return $(".Texto:first",
> > > > > > > > $(this)).text();
> > > > > > > > });
> > > > > > > > var slots2 =
> > > > > > > > $(this).parents('dl').map(function() {
> > > > > > > > return $(".Texto:first",
> > > > > > > > $(this)).text();
> > > > > > > > });
> > > > > > > > return slots.get().join("/") + ' -- ' +
> > > > > > > > slots2.get().join("/");
> > > > > > > > };
>
> > > > > > > > The funny thing of this code is that slots and slots2 have the
> > > > > > > > same
> > > > > > > > items
> > > > > > > > Could you point me why slots put the last dl?
>
> > > > > > > > I suppose that dl:not(:last) retrieves all the parents except
> > > > > > > > the last
> > > > > > > > one but this don't work
>
> > > > > > > > Thanks!
