Throughout this edition, Mark Allen Weiss has included the latest features of the C++ programming language-making prevalent use of the Standard Template Library (STL) wherever appropriate. He has also included a new chapter covering Patterns, as well as new material on templates, vectors and push_back.

C++ allows the programmer to write the interface and implementation separately, to place them in separate files and compile separately, and to hide the implementation details. This book goes a step further: The interface and implementation are discussed in separate parts of the book. Part I (Objects and C++), Part II (Algorithms and Building Blocks), and Part III (Applications) lay the groundwork by discussing basic concepts and tools and providing some practical examples, but implementation of data structures are not shown until Part IV (Implementations). This separation of interface and implementation promotes abstract thinking. Class interfaces are written and used before the implementation is known, forcing the reader to think about the functionality and potential efficiency of the various data structures (e.g. hash tables are written well before the hash table is implemented).

Please excuse me for my poor english, it is not my primary languange.

Mr. Weiss has an impressive academic record. With a PhD from Princeton in Computer Science and having studied under Dr. Sedgewick, a king of algorithms, he now teaches at Florida International University. Clearly, he is an authority on the subject. However, he should pursue something other than writing textbooks. Let's examine the flaws in his approach.

The author works heartily to impress you with his ability to demonstrate proofs for algorithms. He illustrates various proofs throughout the text and gives a liberal discussion of each. He then provides implementations of each algorithm and spends considerable time discussing the C++ language statements that are used. What he never does adequately is to provide a discussion of data structures and algorithms absent the details which distract the reader of an introductory text. As a point of comparison and constrast, I ask you to consult Data Structures with C++ using STL by William Ford/William Topp.

Consider the study of stacks and how compilers utilize them to evaluate expressions. This is a subject taught in every data structures class. Let's examine the approaches of Weiss and Ford/Topp.

Pages devoted to the topic. Weiss ( 38 ) vs Ford/Topp ( 45 )
Number of tables+figures(excluding code listings) Weiss( 5 ) vs Ford/Topp( 27 )
Pages devoted to writing about data structures ( not C++ implementation ) Weiss ( 8 ) vs Ford/Topp ( 17 )
Questions at the end of chapter with solutions provided . Weiss ( 0 ) vs Ford/Topp ( 11 )

From the brief analysis above, you may recognize a pattern that Dr. Weiss himself admits to following in the writing of his book. In the preface, he says that "fundamental choices need to be made including ... level of mathematical rigor, balance between implementation of data structures and their use ... and programming details related to language chosen ". Regrettably Dr. Weiss has made poor choices on all three areas.

First, the mathematical rigor which might be highly valued in a graduate level text is seen here as a mere annoyance. Second, the balance between implementation and their use is skewed toward the former when it is needed toward the latter. Finally, students who have had adequate instruction in C++ do not need to be led by the hand as the author guides the reader through an implementation. Too many other authors do a much better job with C++ than Dr. Weiss ( read Schildt, Meyers, many more ). Dr. Weiss would be well advised to keep the code examples in play but cut out the narratives on the code.

I was able to do well in my class because I recognized the deficiencies in Dr. Weiss' book and quickly abandoned it in favor of the Topp/Ford text.

On a final note of disappointment, the binding of this book, purchased brand new, failed after just two weeks. I treat my books with respect and in the last twenty years have had only one other occasion of a binding failure ( the pages separate from the web and glue that holds everything together ). My bookseller replaced the text without question.

You will probably be required to buy this book for your data structures class ( as I was ). It is a sad waste of money but many faculty remain impressed with Dr. Weiss' output. I recommend buying a used version ( absent the broken bindings ) and gettinf another readable text from a different author. To be fair, I did give this book a 2 rather than a 1 ( worst ) on a scale of 1 to 5. Since I am planning graduate study in computer science, I believe that the book might hold some value as a reference for the very items that I criticize here. That is, the proofs might be of some interest in a advanced data structures or advanced algorithms course.