In the 'Notes to the Reader' section 1.2.1 on page 11 of 'The C++ Programming Language', Bjarne Stroustrup mentions the four major supported styles of programming in C++ -
- Procedural programming
- Data abstraction
- Object-oriented programming
- Generic programming
He later attempts to describe data abstraction in terms of classes, esp. with respect to abstract classes. Unfortunately, he does not draw a clear distinction between classes as used for data abstraction and as used for object-orientation. This is particularly wanting because PLT shows quite clearly that not only is data abstraction different from object-orientation, they are in fact duals. It is a further confounding that the mechanisms used to implement each style have a surprising degree of overlap.
Another overlooked point is how uniquely powerful is the combination of data abstraction and Bjarne's fourth type of style - generic programming. Taken together, you have a foundation for functional programming in C++ (similar to my talk on abstract data types in F# here - https://vimeo.com/128464151).
Unfortunately, for the rest of the book, and for the rest of all C++ literature I've encountered, the concept of data abstraction remains unexplored.
However, as the world wakes up to the broken promises of object-orientation, it is the time to revisit this lesser discussed concept of data abstraction.
In this repository, I give a strong and precise definition of data abstraction, and data-abstraction-style in C++. Additionally, I build both a very small standard library called 'das', which also includes a publisher-neutral event system, as well as a couple of toy demo programs. All of these are implemented in the data abstraction-style. The result has been code that is simpler, more efficient, and more general than code implemented in the object-oriented style.
To get started, begin by reading the first literate tutorial here - https://github.com/bryanedds/das/blob/master/src/hpp/tut/tut.hpp
Have fun, and please report any feeback, criticism, oversights, or errors as a new issue here - https://github.com/bryanedds/das/issues