I want to select nodes which have attribute class with some value not specifying tag.
How to do it?
So far I have:
html.DocumentNode.SelectNodes("//[@class='value']"); But it's not working good as far as I see it.
For instance, let me have this kind of HTML code:
<div> <div class="value"></div> <a class="value"></div> </div> it would need to give me back those 2 elements inside of <div>, so <div> and <a>. Is that possible?
So I believe the correct syntax for what you want is this:
html.DocumentNode.SelectNodes("//*[@class='value']"); When you're looking for just the attribute you don't need the [ ] because you're not refining it by element.
If you want to include the root node, you can use the descendant-or-self access like so:
descendant-or-self::*[@class='value'] Eliminate -or-self if you don't want to consider the root node. More importantly, the asterisk is what is telling the XPath parser to return a node set.
Because you're working with HTML class attributes here, I suggest using the following:
//*[contains(concat(' ', @class, ' '), ' value ')] Note the spaces surrounding the value. This will ensure that you find the elements even if they have multiple classes. Simply using @class='value' would not work in that situation.
