Well, he recommends both of them and he said he will be giving exercises from both of these texts, although I have a feeling that most of these will be from Jackson since it's the standard and there are more copies in the library.
Jackson is the standard source. I like it very much, but Schwinger is simply a master piece, and I like. It's full of gems. Alone the treatment of cylindrical Bessel functions with elegant operator techniques is worth the whole book.
I don't like both books conceptionally wise so much. They are both written from a classical perspective. From a HEP/nuclear theoretical physicist's point of view, classical electromagnetism should be introduced as a relativistic classical quantum field theor (and it's in fact the paradigmatic example of one such theory with the clear advantage that it describes a vast amount of observable facts and has far-reaching applications in electrical engineering, electronics, and optics). The best book from this point of view is, in my opinion, Landau&Lifshitz vol. 2 or Scheck's Theoretical Physics book vol. 3 (I'm not sure whether there's an English translation of the latter, but the German edition is great).
Buy the one most used in class, and if your professor is going to swap about equally, then you probably should buy both. You can buy an older edition of Jackson, it is the standard and older editions are probably readily available for a LOT less money than the newest edition (also look at Indian publishers).
My next advice, buy a study guide if you can and buy an REA or Schaum's workbook. I had a professor that couldn't be bothered with stupid questions (the only kind I knew how to ask...), so that was a necessity for me. We used two textbooks, Jackson and Griffiths. Griffiths was far easier to read and the problems were far easier to do. I suspect Grittiths was really an undergrad text, but it complemented Jacksons and allowed the class to comprehend the more difficult concepts.