I'll point out that there are more <del> elements in your sample than <ins> elements. So if you tried matching everything between <del> and </ins> you'll either have <del> elements left over or you'll be matching stuff that isn't inside a <del> element or an <ins> element. Is either of these what you want?
By "extract" do you mean you want it removed? Because you can load the HTML into a DOM object and then remove those elements. In fact, whatever you're planning to do, working with the HTML document as an HTML document would be the most robust approach.
If the document is really that simple, then you could get away with a regular expression; off the top of my head
/<(del|ins)>(.*?)<\/1>/ will capture the surrounded text and whether it was surrounded by <del> or <ins> tags.