Roark's formulas for stress and strain hadn't satisfied me: information is not oriented for structural engineers, introductory texts are not enough theoretical and you have US units throughout.
In Pilkey's book, you have the perfect structural engineer's reference: many chapters, with at first a list of notation, explanation of conventions, and then a short introductory course on the subject together with solved examples. After that, there it is: magnificent well-organized "tables", with all kind of data of prime interest to a structural engineer. As an example, I'll mention that you can find plastic section modulus for about 11 section types.
Units are mixed for examples, but for data you have always both US and SI units furnished.
For all entries, Pilkey's book is far more complete than the Roark's one. You'll be surprised by the vastness and depth of formulas furnished. Furthermore, you have structural matrices in each case if you want to do numerical programming.
The list of references is up to date and very extensive. It is a pricy book, but you'll not regret it!