Question

If I want to parse a HTML block using Nokogiri in Ruby like this:

<th class="first">ancd</th>
<th>xyz</th>
<th>sdf</th>

How do I exclude the tag containing a certain class name? the “first” in this case.

Answer 1

You can use CSS selectors:

doc.css('th:not(.first)')

In this simple case, you can also use xpath:

doc.xpath('//th[not(@class="first")]')

The difference is that xpath requires an exact match in the class name. If you had the possibility of having multiple classes, e.g. <th class="red first"> then the CSS selector will identify it, but the xpath won’t (without making it a bit more complicated).

Edit: Just for reference, if you want the xpath that can select a class when there may be multiple classes on an element:

doc.xpath('//th[@class and contains(concat(" ",normalize-space(@class)," "), " first ")]')

Normally XPath is a lot more flexible than CSS selectors, but this is an HTML edge case that favors CSS.