Adding support for getting an element's parent

[andyw8]

It seemed Capybara lacked a simple way to get an element's parent ( http://stackoverflow.com/questions/4861863/how-to-get-parent-node-in-capybara ) so I've added support for this.

[jnicklas]

Just do find(:xpath, '..'). No need to expose these kinds of traversal methods in Capybara.

[cj]

@jnicklas is there a way to do that but lazy find the body? I'm trying to get back to the body from inside a within, so that I can check it loaded a popup....

[cj]

@jnicklas looks like find(:xpath, '//body') works :D

[dgm]

How is this not a useful addition? Not everyone is comfortable with xpath... I'd prefer to have a jquery like navigation available. Even more useful would be something like http://api.jquery.com/closest/

[mikebaldry]

Sorry for dragging an old issue up.

I don't agree with jnicklas. capybaras purpose is to allow us to test against a page, which may include finding an element then traversing to another element to check something. It may be slightly less common than click "somebutton" but its still valuable.

[dfl]

+1 I would like this feature too, even if it's just syntactic sugar

[jnicklas]

I'm not against this feature, though it is just syntactic sugar, so I find it a bit redundant. This pull request though is just completely wrong. So if someone wants this, please send a better pull request.

[janko]

What does Capybara::Node::Base#parent attribute currently do? It seems to be always returning Capybara::Document instead of the actual tag's parent.

[twalpole]

@janko-m Its the context for the find which located the current element.

[abotalov]

@janko-m Base#parent is internally used to get an ancestor using which the current element was located. For example in:

el = find('a')
el.find('b').parent # => returns el